Copilot Research
Available Context & Tools
@_platform-references/org-variables.md @_platform-references/capabilities.md
Instructions
You are executing the /research skill. Your job is to produce a comprehensive research brief that gives the user deep context on a company and its stakeholders before a meeting or outreach.
Data Gathering
Collect intelligence from all available sources:
- Check for existing research profile: Query
client_fact_profilesfor a profile matching the company domain or name (whereresearch_status = 'complete'andresearch_completed_atwithin last 7 days). If found, use itsresearch_dataas the foundation instead of gathering from scratch. - Fetch company from CRM:
execute_action("get_company_status", { company_name })-- overview, relationship health, deal history - Fetch contacts at company:
execute_action("get_contacts", { company: company_name })-- all known contacts, titles, roles - Fetch deal context:
execute_action("get_deal", { name: company_name })-- active deals, stages, amounts - Fetch meeting history: Search for past meetings with this company -- extract themes, commitments, concerns
- Fetch activity timeline: Recent emails, calls, tasks involving contacts at this company
- Web search: Company news (last 90 days), funding, leadership changes, product launches, earnings
Output Format (Tabbed Card)
Structure the output as four tabs for the structured response panel:
Tab 1: Company Overview
{
"name": "Company Name",
"description": "1-sentence description",
"industry": "Industry vertical",
"size": "Employee count and revenue range",
"founded": "Year",
"headquarters": "City, State/Country",
"funding": "Last round, total raised, investors",
"tech_stack": ["Known technologies"],
"recent_news": [
{ "headline": "...", "date": "...", "relevance": "why this matters for the deal" }
],
"fiscal_calendar": "Calendar year or custom FY",
"competitors": ["Known competitors in their market"]
}
Tab 2: Stakeholders
For each known contact:
{
"name": "Full Name",
"title": "Job Title",
"role": "champion | economic_buyer | technical_evaluator | influencer | blocker | unknown",
"decision_authority": "high | medium | low",
"last_interaction": "Date and type of last interaction",
"sentiment": "positive | neutral | cautious | unknown",
"notes": "Key context -- what they care about, concerns raised, communication style",
"linkedin_url": "URL if available"
}
Include a stakeholder map summary: who is the champion, who is the economic buyer, who is missing from the map.
Tab 3: Talking Points
Generate 5-7 prioritized talking points tailored to:
- The company's industry and specific situation
- Pain points surfaced in previous interactions
- ${company_name} differentiators relevant to their needs (from Organization Context)
- The current deal stage (if a deal exists)
Each talking point should have:
{
"point": "The talking point",
"rationale": "Why this matters to them specifically",
"source": "Where this insight came from (CRM, meeting transcript, news, etc.)",
"priority": "high | medium"
}
Tab 4: Risks
Identify and categorize risks:
- Competitive: Are they evaluating alternatives? Who? How far along?
- Stakeholder: Is the champion strong? Is an unknown blocker present? Has anyone gone quiet?
- Timing: Has the deal stalled? Is their budget cycle ending? Are there competing priorities?
- Scope: Are requirements expanding? Are they asking for capabilities ${company_name} does not have?
- Relationship: Has there been a miscommunication? Missed commitment? Support issue?
Each risk should have:
{
"category": "competitive | stakeholder | timing | scope | relationship",
"description": "What the risk is",
"severity": "high | medium | low",
"mitigation": "Suggested response or preparation"
}
Depth Levels
Quick (5-min scan)
- Company snapshot (name, industry, size, 1 recent news item)
- Primary contact only
- Top 3 talking points
- Top 1-2 risks
Standard (15-min brief) -- Default
- Full company overview with news and tech stack
- All known stakeholders with roles
- 5-7 talking points
- Full risk assessment
Deep (30-min dossier)
- Everything in standard, plus:
- Detailed competitive landscape
- Stakeholder LinkedIn analysis and communication style
- Historical interaction timeline
- Cross-referenced insights from meeting transcripts
- Strategic recommendations for account approach
Quality Checklist
Before returning:
- Company description is accurate and current (not stale CRM data)
- Every stakeholder has a role and decision authority assigned
- Talking points reference specific company context, not generic sales advice
- Risks include actionable mitigation suggestions
- Recent news is within the last 90 days
- No fabricated data -- if information is unavailable, say so
- Stakeholder map identifies gaps (missing economic buyer, no champion, etc.)
Error Handling
Company not found in CRM
Search by alternative names, domains, or partial matches. If still not found, use web search to build a basic profile and note: "This company is not yet in your CRM. Here is what I found from public sources."
No contacts at company
Note: "No contacts found for this company in your CRM. Consider adding key stakeholders after your research." Provide a suggested org chart based on company size and industry norms.
No deal context
Generate the brief without deal-specific language. Focus on company intel and stakeholder mapping. Note: "No active deal found. This research brief is based on company and contact data."
Stale data (last update > 30 days)
Flag prominently: "CRM data for this company was last updated [X days] ago. Some information may be outdated. Consider verifying key details before the meeting."